[Internal-cg] ICG call
Milton L Mueller
mueller at syr.edu
Tue Feb 24 21:22:07 UTC 2015
> First off, I think it is not the ICG opinion that matters. Once again, we need
> broad community backing of a plan to make it credible. I am more interested
> in seeing that we get results that are backed by the communities and that
> have gone through a sufficient process, rather than inserting my own
> opinions on what the proposal should say. The CWG needs to see
Amen to Jari.
Of course the NTIA has a legitimate concern about terminating the contract before anything solid is in place, and perhaps that is all they mean by "implementation." But the NTIA criteria for the transition are clear and have been in front of us for almost a year now. We translated those criteria into our RFP. These well-established criteria must not be supplanted by the game of "guess what Larry wants." Since "level of implementation" is not a well-defined thing and is not included within those criteria, we have to leave it to the operational communities to specify the level of implementation detail they are capable of and comfortable with. All we need to do is ensure that whatever they specify doesn't risk deviating from the NTIA criteria as it is implemented.
As Jari says, the really crucial gating factor here, both politically and operationally, is what the community can accept and will support. If there is support there will be implementation. It is inconceivable to me that NTIA would reject or turn back a feasible plan, with a reasonable level of implementation detail, that had won broad support from the community. (But if they did, they would pay a very steep price politically, both domestically and in the international arena.)
Maybe I am wrong, and I would be happy to be corrected if I am, but I get the sense that this sudden emphasis on level of implementation is driven by a concern about the timeline. But to me there is a gaping logical hole in this approach: pre-specifying more implementation detail will take _more_ time, not less - unless the demand for implementation is also a demand for doing something quick and dirty at the expense of quality and consensus.
> themselves as being in charge, and they need to make the difficult decisions
> and create community consensus so that the world has something that we
> can rely on wrt future IANA arrangements on names.
> Secondly, that being said I think the answer on what level of implementation
> detail is sufficient is that it depends. Situations are different. As an example,
> 98% of what is in the IETF proposal is something that we've had in place for
> a decade, and we have lots of experience from it. I think that definitely
> qualifies as sufficient detail :-) And the remaining 2% are things where the
> community found the level of guidance in the proposal sufficient. I have no
> worries in this respect.
> But if you have a different situation, you may want a differing level of detail.
> Lets say you have a proposal that creates a completely new entity for some
> task. While requiring a decade of experience on that new thing might be
> going too far :-) I think some level of detail is needed. Do you need all
> contracts everywhere, ready to be signed? No. Do you need to have a clear
> idea of what you are creating? Yes. For a new entity, a charter would
> definitely be in my list of ready things to have.
More information about the Internal-cg